Optimal Reinsurance Maximising Dividends: An Infinite-Dimensional Optimisation Approach and Numerical Results

49 Pages Posted: 19 Dec 2023

See all articles by Debora Daniela Escobar

Debora Daniela Escobar

London School of Economics and Political Science, Department of Statistics

Hirbod Assa

University of Essex - Department of Mathematics

Yunzhou Chen

University of Liverpool

Abstract

This study considers a discrete-time dynamic framework for designing optimal reinsurance contracts to maximize the lifetime dividends of an insurance company. First, we address the problem under a budget constraint and subsequently incorporate a solvency condition for the insurer. Surprisingly, both settings prove equivalent to solving static problems. Our results emphasize the necessity of a tailored dividend rule to accommodate the dynamic nature of the setup. Specifying a dividend rule adds complexity to the problem, whose solution requires solving an infinite-dimensional Lagrangian problem. We show that among all admissible reinsurance contracts, the optimal solutions for all problems are multi-layer contracts. However, the layered structure from the latter problem depends on Lagrangian multipliers. Despite this, we propose a linear programming problem to approximate its solution, particularly leveraging the barrier dividend rule. Numerical results are presented for the Value-at-Risk (VaR), Average Value-at-Risk (AVaR) and Glue Value-at-Risk (GlueVaR), with a claim that these findings are applicable to any distortion risk measure. The study not only provides theoretical insights into the optimal reinsurance contract design but also offers a practical linear programming algorithm for solution approximation in real-world scenarios.

Keywords: Optimal reinsurance, Distortion risk measures, Dividends, Dynamic problem, Linear programming

Suggested Citation

Escobar, Debora Daniela and Assa, Hirbod and Chen, Yunzhou, Optimal Reinsurance Maximising Dividends: An Infinite-Dimensional Optimisation Approach and Numerical Results. Available at SSRN: https://ssrn.com/abstract=4669559 or http://dx.doi.org/10.2139/ssrn.4669559

Debora Daniela Escobar (Contact Author)

London School of Economics and Political Science, Department of Statistics ( email )

London
United Kingdom

Hirbod Assa

University of Essex - Department of Mathematics ( email )

Wivenhoe Park
Colchester, Essex CO4 3SQ
United Kingdom

Yunzhou Chen

University of Liverpool ( email )

Chatham Street
Brownlow Hill
Liverpool, L69 7ZA
United Kingdom

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
39
Abstract Views
141
PlumX Metrics